Skip to main content

Scientific Analysis Reveals Vincent van Gogh’s True Colors

Scientific analysis of several paintings has revealed new information about Vincent van Gogh’s use of color, an unattributed work believed to be by Spanish master Diego Velazquez, and the materials used by artists in the Roman era, according to conservation experts at a 14 February news briefing at the 2016 AAAS Annual Meeting.

“Science plays a big role in deepening our understanding of these iconic masterpieces,” said Francesca Casadio, A.W. Melon senior conservation scientist at the Art Institute of Chicago, which is hosting a new exhibition, Van Gogh’s Bedrooms. The exhibit reunites three versions of Bedroom at Arles painted by Vincent van Gogh in 1888 and 1889. It also presents the results of several years of collaborative scientific study of the paintings.

Casadio noted that letters from van Gogh precisely describe the colors in the series of paintings: yellow furniture, red bedspread, purple doors, and the walls a pale lilac, which are intended to convey a restful air.

Yet, one version of the painting clearly shows icy blue walls, said Casadio, holding up a print of the version housed in the Art Institute’s permanent collection.

What colors did van Gogh truly intend for the bedroom walls? Casadio’s team used a variety of techniques to determine the palettes, including conventional methods like investigation with stereo microscopes as well as more sophisticated techniques like micro-X-ray fluorescence. They then removed a small sample of paint from the painting’s blue walls, which revealed strata of color “exposed like a layer cake,” Casadio said.


Francesca Casadio displays a Vincent van Gogh painting as it appears today, right, and as it would have looked in the past, left. | Boston Atlantic Photography

“Deep within the stratography of the blue paint, there were remaining particles of pink,” made from carmine lake, Casadio said. The naturally derived, light-sensitive pigment had faded, with only traces remaining under the blue pigment. “Of course, pink and blue makes purple,” she said.

Traces of pink carmine lake were also found in the other two paintings in the series, Casadio said.

To visualize what the colors may have actually looked like when van Gogh painted them, Casadio’s team worked with a color scientist using algorithms to create a “virtual knob that we can use and turn to bring the purple back into the bedroom,” she said, displaying the results of the re-colorized bedroom.

“We’re not only interested in process and technique but also in understanding the emotional landscape that van Gogh was going through as he painted this work,” Casadio said.

Scientific examinations have also provided insight into materials and techniques of a 17th century painting long thought to be by an anonymous Spanish artist, bolstering evidence that the painting could be an early work by Diego Velazquez, according to Ian McClure, the Susan Morse Hilles chief conservator at the Yale University Art Gallery.

The painting had been in Yale’s collection since the 1920s, but it was not until the 2000s that another curator noticed the painting in storage in poor condition and thought it could be by Velazquez, McClure said.

McClure’s examination revealed that the painting, known as Education of the Virgin, was not a copy, despite its strong resemblance to a painting by Juan de Roelas with a similar subject: St. Anne and St. Joachim teaching their daughter, the Virgin Mary, to read. Microscopic examination of an X-ray shows that the artist spent a long time rearranging the composition, he said. Additionally, fine white lines revealed by the X-ray show that the painter was drawing precise perspectival shapes, which has served as “a guide to understand some of the techniques Velazquez used,” McClure said.

Researchers like Aniko Bezur also looked at indicators to tie the painting to Velazquez’s home of 17th century Seville, including ground material covering the canvas surface onto which the artist painted, said Bezur, Wallace S. Wilson director of the Technical Studies Lab at Yale University’s Institute for the Preservation of Cultural Heritage.

The pigment palette used was consistent with that used by Velazquez and his contemporaries in 17th century Seville, she said, as was the material combination practices like the use of certain pigments together and the use of paint to create volume on the paintings. Their research also identified the specific ground material as related to that documented on early Velazquez paintings, Bezur said.

“We can never say for sure that this painting is by Velazquez, but all the evidence that we’ve gathered, both from the point of view of art history and the scientific evidence, all points to it being by Velazquez,” McClure said.


Ian McClure, right, displays a Juan de Roelas painting in a press conference during the 2016 AAAS Annual Meeting.  | Boston Atlantic Photography

Analysis of pigments has also shed light on a series of portraits created during the Roman period in Egypt more than 2,000 years ago.

The team of Marc Walton, research associate professor of materials science and engineering at Northwestern University, used techniques like hyperspectral imaging and photometric stereo, which reveals the surface shape of objects. Based on the data they gathered, they then took very small samples to determine pigment contents, which revealed a surprising conclusion about the materials.

“Not all of them come from the local environment in Egypt,” Walton said. “They easily could have been painted with materials locally available, but they deliberately chose not to do this.”

The painters used materials such as a red lead that came from as far away as Spain, iron-earth pigment from Greece, and wood substrates—onto which the portraits were painted—from central Europe, he said.

“We have more of an international picture, not only of trade during the Roman period, but the outlook of these particular individuals,” Walton said.

The subjects, however, haven’t been identified. Although papyri exist to describe who they are—mostly priests from the cult of Sobek—they were removed during excavation more than a century ago and no longer correspond with the correct paintings, Walton said.